Search for a command to run...
Subducting topographic features—including seamounts and ridges—introduce frictional complexities that can affect megathrust seismicity. Whether these heterogeneities tend to facilitate large earthquakes by acting as asperities or impede rupture propagation as barriers remains debated. To address this, we combine laboratory stick-slip experiments with quasi-dynamic numerical simulations to investigate how frictional barriers affect earthquake cycles. The experiments show that increasing the area of a single circular barrier within an asperity leads to a transition from periodic stick-slip (seismic events) to aseismic slip when the barrier area exceeds a threshold between 8 and 11% of the apparent contact area of the sample. For multiple barriers, the orientation relative to the loading direction strongly influences slip behaviour, with linear arrangements perpendicular to loading promoting a mix of seismic and aseismic slip, while parallel arrangements predominantly exhibit aseismic behaviour. Numerical simulations of a synthetic segmented fault and the Alaska-Aleutian subduction zone show that a velocity-strengthening patch acts as a permanent seismic barrier when its along-strike length constitutes 0.38–0.43 of the fault segment. This range aligns with—and extends slightly beyond—the threshold from our laboratory experiments, where the dimensionless ratio Lb/L falls between 0.32 and 0.38. In contrast, the subsurface ridges along the Himalayan deformation front are too small to act as permanent barriers, allowing dynamic ruptures to propagate across them under favourable conditions. The consistency of this dimensionless parameter across methods provides a robust metric for evaluating barrier potential in subduction zones. Our results align with previous dynamic rupture models, which show that stress transfer and fault rheology facilitate cascading ruptures. While paleoseismic data provide snapshots of past events, numerical modelling reveals physically plausible scenarios beyond current empirical constraints. Collectively, these findings suggest that the role of frictional barriers in modulating earthquake behaviour depends not only on their size but also on their geometry and spatial distribution relative to the plate velocity. These findings have significant implications for understanding objective seismic hazards in subduction zones characterised by frictionally heterogeneous plate interfaces.